CN106791222A - The method and apparatus for recognizing reentry network users - Google Patents

The method and apparatus for recognizing reentry network users Download PDF

Info

Publication number
CN106791222A
CN106791222A CN201611117570.4A CN201611117570A CN106791222A CN 106791222 A CN106791222 A CN 106791222A CN 201611117570 A CN201611117570 A CN 201611117570A CN 106791222 A CN106791222 A CN 106791222A
Authority
CN
China
Prior art keywords
opposite
call
calling
new
fingerprint
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN201611117570.4A
Other languages
Chinese (zh)
Inventor
陈少权
杜翠凤
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
GCI Science and Technology Co Ltd
Original Assignee
GCI Science and Technology Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by GCI Science and Technology Co Ltd filed Critical GCI Science and Technology Co Ltd
Priority to CN201611117570.4A priority Critical patent/CN106791222A/en
Publication of CN106791222A publication Critical patent/CN106791222A/en
Pending legal-status Critical Current

Links

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04M—TELEPHONIC COMMUNICATION
    • H04M3/00—Automatic or semi-automatic exchanges
    • H04M3/22—Arrangements for supervision, monitoring or testing
    • H04M3/2218—Call detail recording
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04M—TELEPHONIC COMMUNICATION
    • H04M3/00—Automatic or semi-automatic exchanges
    • H04M3/22—Arrangements for supervision, monitoring or testing
    • H04M3/229—Wire identification arrangements; Number assignment determination
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W24/00—Supervisory, monitoring or testing arrangements
    • H04W24/08—Testing, supervising or monitoring using real traffic

Landscapes

  • Engineering & Computer Science (AREA)
  • Signal Processing (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Collating Specific Patterns (AREA)

Abstract

Method and apparatus the present invention relates to recognize reentry network users.Methods described includes：The detailed list information of calling of new/old number is obtained, the calling fingerprint storehouse of new/old number is set up respectively；Packet in the calling fingerprint storehouse contains：The call characteristic information of opposite-terminal number and each opposite-terminal number；The call characteristic information of opposite-terminal number includes：Its talk times and the duration of call with the new/old number, and itself and institute's number talk times summation and duration of call summation；Call characteristic information according to each opposite-terminal number calculates the influence coefficient of each opposite-terminal number in the calling fingerprint storehouse of new/old number；Effective call fingerprint base is drawn from the calling fingerprint storehouse of new/old number according to influence coefficient；The similarity of effective call fingerprint base according to new digit and the effective call fingerprint base of old number, judges whether the new digit attaches most importance to networking user.The present invention can improve the degree of accuracy of counterweight networking user identification.

Description

The method and apparatus for recognizing reentry network users
Technical field
The present invention relates to mobile communication technology field, more particularly to a kind of method and apparatus for recognizing reentry network users.
Background technology
In recent years, with the development of mobile network's technology, the atrophy in landline telephone market, mobile telephone service development, fortune Battalion business has welcome many new challenges, such as network re-entry phenomenon.
Network re-entry phenomenon refers to mobile communication subscriber and is possessing one or possessing the feelings of multiple mobile service numbers Under condition, the new digit of institute's home-operator being bought again and being networked, new digit will in whole or in part substitute original old number Business is used.Network re-entry phenomenon is usually expressed as high networking rate, while user's churn rate of operator also remains high, makes Obtain Operator Specific Service development and " importing and exporting on a big scale " phenomenon occur.
Network re-entry causes the card number wasting of resources, market to increase emptily the strategic cloth that high consequence high degree have impact on operator Office, however it is relatively low for the recognition accuracy of reentry network users at present, cause operator to lack and make science decision and specific aim The foundation that marketing is sold.
The content of the invention
Based on this, the method and apparatus for recognizing reentry network users are the embodiment of the invention provides, it is possible to increase to network re-entry The degree of accuracy of user's identification.
One aspect of the present invention provides a kind of method for recognizing reentry network users, including：
The detailed list information of calling of new/old number is obtained, list information sets up exhaling for newly/old number respectively in detail according to the calling It is fingerprint base；Packet in the calling fingerprint storehouse contains：All opposite-terminal numbers conversed with new/old number, and respectively The call characteristic information of opposite-terminal number；The call characteristic information of each opposite-terminal number includes：Its call time with the new/old number Number and the duration of call, and itself and institute's number talk times summation and duration of call summation；
According to the call characteristic information of each opposite-terminal number, the calling fingerprint storehouse Zhong Ge opposite ends number of new/old number are calculated respectively The influence coefficient of code；Influence coefficient according to each opposite-terminal number filters out infecctive call from the calling fingerprint storehouse of new/old number Fingerprint, draws effective call fingerprint base；
Effective call fingerprint base and the similarity of the effective call fingerprint base of old number according to new digit, judge described new Whether number attaches most importance to networking user.
Another aspect of the present invention provides a kind of device for recognizing reentry network users, including：
Calling fingerprint acquisition module, the detailed list information of calling for obtaining new/old number, according to the detailed list information of the calling The calling fingerprint storehouse of new/old number is set up respectively；Packet in the calling fingerprint storehouse contains：It is all with described new/old number The opposite-terminal number of call, and call characteristic information；The call characteristic information of each opposite-terminal number includes：It is new/old with described The talk times and the duration of call of number, and itself and institute's number talk times summation and duration of call summation；
Fingerprint filtering module, for the call characteristic information according to each opposite-terminal number, calculates the calling of new/old number respectively The influence coefficient of each opposite-terminal number in fingerprint base；Coefficient from the calling fingerprint storehouse of new/old number is influenceed according to each opposite-terminal number In filter out infecctive call fingerprint, draw effective call fingerprint base；
Network re-entry determination module, for the effective call fingerprint base of the effective call fingerprint base according to new digit and old number Similarity, judge whether the new digit attaches most importance to networking user.
Above-mentioned technical proposal, the detailed list information of the calling based on new/old number, sets up the calling fingerprint storehouse of new/old number；And Talk times and the duration of call in calling fingerprint information carry out denoising to the fingerprint in initial calling fingerprint storehouse, filtering Fall invalid call information, obtain effective calling fingerprint information, carrying out reentry network users based on effective calling fingerprint storehouse sentences It is fixed, can effectively identify that whether new digit attaches most importance to networking user, improves the degree of accuracy of follow-up network re-entry user's identification.
Brief description of the drawings
Fig. 1 is the indicative flowchart of the method for the identification reentry network users of an embodiment；
Fig. 2 is the schematic diagram of the device of the identification reentry network users of an embodiment.
Specific embodiment
In order to make the purpose , technical scheme and advantage of the present invention be clearer, it is right below in conjunction with drawings and Examples The present invention is further elaborated.It should be appreciated that the specific embodiments described herein are merely illustrative of the present invention, and It is not used in the restriction present invention.
Fig. 1 is the indicative flowchart of the method for the identification reentry network users of an embodiment；As shown in figure 1, the present embodiment In identification reentry network users method include step：
S11, obtains the detailed list information of calling of new digit, and the calling fingerprint of new digit is set up according to the detailed list information of the calling Storehouse；The detailed list information of calling of old number is obtained, the calling fingerprint storehouse of old number is set up according to the detailed list information of the calling.
It is understood that in the present embodiment, the when network entry time that new digit refers to is discontented with the number of setting duration, old number Refer to the number that network entry time exceedes setting duration.In the present embodiment, the packet in the calling fingerprint storehouse of new digit contains：Institute Have with the new digit converse opposite-terminal number, and each opposite-terminal number call characteristic information；The calling of each opposite-terminal number Characteristic information includes：Its talk times and the duration of call with the new digit, and it is total with institute number talk times With with duration of call summation.Packet in the calling fingerprint storehouse of old number contains：All opposite ends number conversed with the old number Code, and each opposite-terminal number call characteristic information；The call characteristic information of each opposite-terminal number includes：Itself and the old number Talk times and the duration of call, and itself and institute's number talk times summation and duration of call summation.
For example by the detailed list information of calling, find what is conversed with a number A (new digit or old number) in setting time Opposite-terminal number has 50, respectively 1~number of number 50；So can obtain the packet in the corresponding calling fingerprint storehouses of number A Include 50 finger print informations, respectively 1~number of number 50 and the corresponding call characteristic information of 1~number of number 50.I.e. each Opposite-terminal number one calling fingerprint of correspondence, and in the corresponding calling fingerprint of each opposite-terminal number, not only comprising the opposite-terminal number with The talk times in setting time and the duration of call of the number A, also including the opposite-terminal number in the setting time With other number talk times summations and duration of call summation.
S12, in the calling fingerprint storehouse of new digit, according to the call characteristic information of each opposite-terminal number, calculates each opposite end number The influence coefficient of code；Influence coefficient according to each opposite-terminal number filters out infecctive call and refers to from the calling fingerprint storehouse of new digit Line, draws the effective call fingerprint base of new digit；Also, in the calling fingerprint storehouse of old number, according to exhaling for each opposite-terminal number Characteristic information is, the influence coefficient of each opposite-terminal number is calculated；The calling of influence coefficient from old number according to each opposite-terminal number refers to Infecctive call fingerprint is filtered out in line storehouse, the effective call fingerprint base of old number is drawn.
It is the call in setting time respectively according to 1~number of number 50 and the number A based on the example above Number of times and the duration of call, and 1~number of number 50 in the setting time with other number talk times summations With duration of call summation, the respective influence coefficient of 1~number of number 50 in the calling fingerprint storehouse of number A is calculated.
In a preferred embodiment, the mistake from the calling fingerprint storehouse of new/old number of the influence coefficient according to each opposite-terminal number Infecctive call fingerprint is filtered, drawing the mode of effective call fingerprint base can be：In the calling fingerprint storehouse of new digit, by each opposite end Number and its call characteristic information are ranked up from high to low according to influence coefficient, obtain the preceding setting number of sequence (for example 20) opposite-terminal number and its call characteristic information, draw the effective call fingerprint base of new digit.In the calling fingerprint of old number In storehouse, each opposite-terminal number and its call characteristic information are ranked up from high to low according to influence coefficient, obtain sequence preceding The opposite-terminal number and its call characteristic information of setting number (such as 20), draw the effective call fingerprint base of old number. Follow-up for convenience to calculate, the quantity of the effective finger print information in the effective call fingerprint base of new/old number is equal.
In the present embodiment, talk times and the duration of call based on opposite-terminal number with the new/old number, and itself and institute Number talk times summation and duration of call summation determine the influence coefficient of each opposite-terminal number, that is, determine each opposite end number Importance of the code in calling fingerprint storehouse, calling cohesion that can be between effecting reaction number can fully take into account public again The disturbing factor of the social calling of the non-genuine such as phone, advertisement marketing phone, the influence coefficient for thereby determining that can more accurately reflect Go out the true calling friend-making circle of new/old number, for follow-up network re-entry user's identification provides effective data basis.
S13, effective call fingerprint base and the similarity of the effective call fingerprint base of old number according to new digit, judges institute Whether new digit is stated to attach most importance to networking user.
In a preferred embodiment, can by cosine similarity algorithm calculate new digit effective call fingerprint base with old number The similarity of the effective call fingerprint base of code, if the similarity can determine that the new digit in the number range of setting Attach most importance to networking user.
Cosine similarity algorithm is that individual achievement data is mapped into vector space, then by measure two individualities to Inner product space included angle cosine value between amount measures the similitude between them.In general, the cosine value of angle is bigger, says The similarity of bright two individualities is higher, and vice versa.Two cosine values of vector are calculated using Euclidian dot product formula：
Knowable to formula (1), the similarity of vectorial a and b can be weighed from cos θ.The algorithm can equally weigh multi-C vector it Between similitude.
The method of the identification reentry network users of above-described embodiment, the list information in detail of the calling based on new/old number, set up newly/ The calling fingerprint storehouse of old number；And the talk times and the duration of call in calling fingerprint information are to initial calling fingerprint storehouse In fingerprint carry out denoising, filter out invalid call information, extract effective calling fingerprint information, improve subsequent user knowledge Other degree of accuracy, can effectively identify whether new digit attaches most importance to networking user.
TF-IDF (term frequency-inverse document frequency) algorithm is a kind of statistic algorithm, It is used to assess significance level of the keyword for a copy of it document in a document sets.The importance of keyword is with it The number of times for occurring in a document is directly proportional increase, but under can be inversely proportional with the frequency that it occurs in whole document sets simultaneously Drop.The main thought of TF-IDF algorithms is：If the frequency that certain keyword occurs in a document is high, its TF (Term Frequency, word frequency) it is high；And seldom occur in other documents of document sets, then it is assumed that the keyword has good class Other separating capacity, is adapted to classification, i.e., its IDF (Inverse Document Frequency, reverse document-frequency) is also high. Wherein, the computing formula of TF values is：
In formula (2), tfiRepresent word frequency of the keyword in document i；Molecule is the frequency that keyword occurs in document i Rate, denominator is that the frequency occurred in whole k documents of the keyword adds up to.
Wherein, the size of IDF values represents the measurement of certain keyword general importance, can by total number of documents divided by comprising The number of the document of the keyword, then the business that will be obtained takes the logarithm and obtains, and its computing formula can be：
In formula (3), idfiRepresent reverse document-frequency of the keyword in document i；Molecule is document in document sets Quantity；tiRepresent certain keyword t, denominator { d included in document i:ti∈ d } represent document sets in include keyword tiDocument d Quantity, if in document sets include keyword tiDocument d number be 1, then it represents that in document sets only have document i in include The keyword.
Based on formula (2) and (3), TF-IDF is defined as：
TF-IDFi=tfi*idfi； (4)
It can thus be concluded that arriving keyword tiTF-IDF statistical values in document i.
In a preferred embodiment, calculated using improved TF-IDF algorithms in the calling fingerprint storehouse of new/old number respectively The influence coefficient of opposite-terminal number, concrete mode is：
In formula (5), TF-IDFuvIt is the shadow of opposite-terminal number v in the calling fingerprint storehouse of number u (new digit or old number) Ring coefficient (i.e. TF-IDF statistical values)；TuvIt is opposite-terminal number v and the duration of calls of the number u in setting time；TvIt is opposite end number Code v in setting time with institute's number duration of call summation；tfuvIt is opposite-terminal number v and number u logical in setting time Words number of times (equivalent to the word frequency in traditional TF-IDF algorithms)；idfuvFor opposite-terminal number v in setting time with institute's number Talk times summation (equivalent to the reverse document-frequency in traditional TF-IDF algorithms).
By the corresponding influence system of each finger print information in above-mentioned improved TF-IDF algorithms calculating original call fingerprint base Number, and then effective calling fingerprint is extracted based on the influence coefficient, have the advantage that：
With the increase of opposite-terminal number v occurrence number and duration of call in the detailed list of calling of number u, i.e., two numbers The common opposite end number of call features increase, in this way, influence coefficients of the number v in the calling fingerprint storehouse of number u also increases therewith Greatly；But, if more to the number of times that calculation number v occurs in the detailed list of all-calls, then number v refers in the calling of number u Corresponding IDF values will reduce in line storehouse, and influence coefficients of such number v in the calling fingerprint storehouse of number u also reduces therewith, It will also reduce to distinguishing the influence power of number u.The frequency that such as some public numbers or sales calls are used is very high, then Its call features in the calling fingerprint storehouse of number u is just weaker.Therefore, the present embodiment uses improved TF-IDF algorithms one Determine to reduce interference of the non-genuine social activity call number to differentiating the degree of accuracy in degree, provided to set up effective calling fingerprint storehouse Important guarantee.
In a preferred embodiment, the effective call fingerprint base of the effective call fingerprint base according to new digit and old number Similarity, judge that whether the attach most importance to mode of networking user of the new digit includes：In the effective call fingerprint base of new digit, First determine the call features weight of each opposite-terminal number respectively according to the influence coefficient of each opposite-terminal number, obtain adding for the new digit Power calling fingerprint storehouse；In the effective call fingerprint base of old number, first determined respectively respectively according to the influence coefficient of each opposite-terminal number The call features weight of opposite-terminal number, obtains the weighting calling fingerprint storehouse of the old number；Then the weighting for calculating new digit is exhaled The similarity of fingerprint base and the weighting calling fingerprint storehouse of old number is, judges whether the new digit attaches most importance to according to the similarity Networking user.With simplified operation, identification validity is further improved.
In a preferred embodiment, in the effective call fingerprint base of new/old number, each opposite end is calculated according to equation below The call features weight of number：
In formula (6), wuiRepresent i-th call features weight of opposite-terminal number in the effective call fingerprint base of number u； TF-IDFuiRepresent i-th influence coefficient of opposite-terminal number in the effective call fingerprint base of number u；N is the effective call of number u The sum of opposite-terminal number in fingerprint base；TuiIt is i-th opposite-terminal number and the duration of calls of the number u in setting time；TiIt is i-th Individual opposite-terminal number in setting time with institute's number duration of call summation；tfuiFor i-th opposite-terminal number is setting with number u Talk times in fixing time, idfuiFor i-th opposite-terminal number in setting time with institute's number talk times summation.
Further, the effective call fingerprint base of new digit and effectively exhaling for old number are calculated using weighting cosine similarity The similarity of fingerprint base is, i.e., the weighting calling in weighting calling fingerprint storehouse and old number that new digit is calculated using equation below is referred to The similarity in line storehouse：
Wherein, Sim (D1,D2) represent number 1 weighting calling fingerprint storehouse D1With the weighting calling fingerprint storehouse D of number 22's Similarity；w1kIt is the weighting calling fingerprint storehouse D of number 11In k-th opposite-terminal number call features weight；w2kIt is number 2 Weighting calling fingerprint storehouse D2In k-th opposite-terminal number call features weight；N is opposite-terminal number in effective call fingerprint base Sum, D1And D2It is n-dimensional vector.In order to shorten the time of calculating, preceding 20 opposite-terminal numbers of sequence and its call features are taken Information structure effective call fingerprint base participates in calculating, i.e. vector D1And D2Dimension be 20.
Recognition methods based on above-described embodiment, to the effective call fingerprint base of the new digit of multiple sampling 30,000 and old number Cosine similarity calculating is carried out, matching result is：
Similarity is interval Number of users Match correct user Accuracy rate
>=0.2419 30000 26451 88.17%
>=0.3802 30000 26619 89.73%
>=0.4537 30000 27468 91.56%
>=0.5536 30000 27951 93.17%
As can be known from the above table, the zone of reasonableness of the number range of the old and new's number similarity is in (0.4537,0.5536).Therefore, Operator can carry out the analysis of network re-entry reason according to the scope to the new user of network re-entry, be that follow-up user safeguards and seeks Pin work provides data supporting.
The method of the identification reentry network users based on above-described embodiment, the technical advantage with following several respects：
(1) data acquisition facility, because the detailed list information of the calling of newly/old number is the information based on operator, it is easy to Obtained by operator；
(2) dynamic of user's interaction network is considered, user's use habit feature of fitting.
(3) computation complexity is not high；Effective relationship cycle is built using improved TD-IDF algorithms, algorithm only considers to use The topological structure factor of family relationship cycle, eliminates tradition and considers the trouble that various social factors are modeled.
(4) recognition accuracy is high；The calling fingerprint storehouse of user is carried out by improved TD-IDF algorithms first it is dry, can To be filled into most invalid calling fingerprint, the error that reduction is recognized below；Secondly carried out with reference to cosine similarity algorithm The identification of reentry network users, can quickly recognize reentry network users.
It should be noted that for foregoing each method embodiment, in order to simplicity is described, it is all expressed as a series of Combination of actions, but those skilled in the art should know, and the present invention is not limited by described sequence of movement, because according to According to the present invention, some steps can sequentially or simultaneously be carried out using other.Additionally, also any group can be carried out to above-described embodiment Close, obtain other embodiments.
Based on above-described embodiment in identification reentry network users method identical thought, the present invention also provide identification weight The device of networking user, the device can be used for the method for performing above-mentioned identification reentry network users.For convenience of description, identification is reentried In the structural representation of the device embodiment of network users, the part related to the embodiment of the present invention, this area skill are illustrate only Art personnel are appreciated that the restriction of schematic structure not structure twin installation, can include than illustrating more or less part, or Person combines some parts, or different part arrangements.
Fig. 2 is the schematic diagram of the device of the identification reentry network users of one embodiment of the invention；As shown in Fig. 2 this The device of the identification reentry network users of embodiment includes：Calling fingerprint acquisition module 210, fingerprint filtering module 220 and reentry Net determination module 230, details are as follows for each module：
The calling fingerprint acquisition module 210, the detailed list information of calling for obtaining new/old number, according to the calling List information sets up the calling fingerprint storehouse of newly/old number respectively in detail；Packet in the calling fingerprint storehouse contains：It is all with it is described The opposite-terminal number of newly/old number call, and call characteristic information；The call characteristic information of each opposite-terminal number includes：Its with The talk times and the duration of call of described new/old number, and it is total with institute's number talk times summation and the duration of call With；
The fingerprint filtering module 220, for the call characteristic information according to each opposite-terminal number, calculate respectively it is new/old number The influence coefficient of each opposite-terminal number in the calling fingerprint storehouse of code；Influence coefficient exhaling from new/old number according to each opposite-terminal number Cry and filter out infecctive call fingerprint in fingerprint base, draw effective call fingerprint base；
The network re-entry determination module 230, exhales for the effective call fingerprint base according to new digit with the effective of old number It is the similarity of fingerprint base, judges whether the new digit attaches most importance to networking user.
In a preferred embodiment, the fingerprint filtering module 220 includes：
Influence coefficient calculation unit, the calling fingerprint storehouse Zhong Ge opposite ends number for calculating new/old number according to equation below The influence coefficient of code：
Wherein, TF-IDFuvIt is the influence coefficient of opposite-terminal number v in the calling fingerprint storehouse of number u；TuvFor opposite-terminal number v with The duration of calls of the number u in setting time；TvFor opposite-terminal number v is total with the institute number duration of call in setting time With；tfuvIt is opposite-terminal number v and talk times of the number u in setting time；idfuvFor opposite-terminal number v in setting time with Institute's number talk times summation；
And/or,
Sequence and screening unit, for the opposite-terminal number and its call features in the calling fingerprint storehouse of new/old number to be believed Breath is ranked up from high to low according to influence coefficient, obtains the opposite-terminal number and its call features letter of the preceding setting number of sequence Breath, draws effective call fingerprint base.
In a preferred embodiment, the network re-entry determination module 230 includes：
Weighting processing unit, in the effective call fingerprint base of new/old number, the influence system according to each opposite-terminal number Number, determines the call features weight of each opposite-terminal number respectively, obtains the weighting calling fingerprint storehouse of the new/old number.Similarity Computing unit, the similarity for calculating the weighting calling fingerprint storehouse of new digit and the weighting calling fingerprint storehouse of old number.And, Identifying unit, for judging whether the new digit attaches most importance to networking user according to the similarity.
It should be noted that above-mentioned example identification reentry network users device implementation method in, each module/unit it Between the content such as information exchange, implementation procedure, due to being based on same design, its skill brought with preceding method embodiment of the present invention Art effect is identical with preceding method embodiment of the present invention, and particular content can be found in the narration in the inventive method embodiment, herein Repeat no more.
Additionally, in the implementation method of the device of the identification reentry network users of above-mentioned example, the logic of each functional module/unit Division is merely illustrative of, in practical application can as needed, such as configuration requirement or software for corresponding hardware The convenient consideration of realization, above-mentioned functions distribution is completed by different functional modules/unit, will the identification reentry network users The internal structure of device be divided into different functional modules, to complete all or part of function described above.It is wherein each Function mould can both be realized in the form of hardware, it would however also be possible to employ the form of software function module is realized.
It will appreciated by the skilled person that all or part of flow in realizing above-described embodiment method, being can Completed with instructing the hardware of correlation by computer program, described program can be stored in embodied on computer readable storage and be situated between In matter, as independent production marketing or use.Described program upon execution, can perform as above-mentioned each method embodiment it is complete Portion or part steps.Wherein, described storage medium can be magnetic disc, CD, read-only memory (Read-Only Memory, ROM) or random access memory (Random Access Memory, RAM) etc..
In the above-described embodiments, the description to each embodiment all emphasizes particularly on different fields, and does not have the portion described in detail in certain embodiment Point, may refer to the associated description of other embodiments.
Embodiment described above only expresses several embodiments of the invention, it is impossible to be interpreted as to the scope of the claims of the present invention Limitation.It should be pointed out that for the person of ordinary skill of the art, without departing from the inventive concept of the premise, Various modifications and improvements can be made, these belong to protection scope of the present invention.Therefore, the protection domain of patent of the present invention Should be determined by the appended claims.

Claims (10)

1. it is a kind of recognize reentry network users method, it is characterised in that including：
The detailed list information of calling of newly/old number is obtained, is referred to according to the calling calling that single information sets up newly/old number respectively in detail Line storehouse；Packet in the calling fingerprint storehouse contains：All opposite-terminal numbers conversed with new/old number, and each opposite end The call characteristic information of number；The call characteristic information of each opposite-terminal number includes：Its talk times with described new/old number and The duration of call, and itself and institute's number talk times summation and duration of call summation；
According to the call characteristic information of each opposite-terminal number, each opposite-terminal number in the calling fingerprint storehouse of new/old number is calculated respectively Influence coefficient；Influence coefficient according to each opposite-terminal number filters out infecctive call fingerprint from the calling fingerprint storehouse of new/old number, Draw effective call fingerprint base；
Effective call fingerprint base and the similarity of the effective call fingerprint base of old number according to new digit, judge the new digit Whether attach most importance to networking user.
2. the method for identification reentry network users according to claim 1, it is characterised in that according to equation below calculate it is new/ The influence coefficient of each opposite-terminal number in the calling fingerprint storehouse of old number：
T F - IDF u v = T u v T v * tf u v * idf u v ;
Wherein, TF-IDFuvIt is the influence coefficient of opposite-terminal number v in the calling fingerprint storehouse of number u；TuvIt is opposite-terminal number v and number u The duration of call in setting time；TvFor opposite-terminal number v in setting time with institute's number duration of call summation；tfuv It is opposite-terminal number v and talk times of the number u in setting time；idfuvFor opposite-terminal number v in setting time with all numbers The talk times summation of code.
3. it is according to claim 1 identification reentry network users method, it is characterised in that according to the influence of each opposite-terminal number Coefficient filters out infecctive call fingerprint from the calling fingerprint storehouse of new/old number, draws effective call fingerprint base, including：
Opposite-terminal number and its call characteristic information in the calling fingerprint storehouse of new/old number is arranged from high to low according to influence coefficient Sequence, obtains the opposite-terminal number and its call characteristic information of the preceding setting number of sequence, obtains effective call fingerprint base.
4. it is according to claim 1 identification reentry network users method, it is characterised in that according to the effective call of new digit The similarity of fingerprint base and the effective call fingerprint base of old number, judges whether the new digit attaches most importance to networking user, including：
In the effective call fingerprint base of new/old number, the influence coefficient according to each opposite-terminal number determines each opposite-terminal number respectively Call features weight, obtain the weighting calling fingerprint storehouse of described new/old number；
The similarity in the weighting calling fingerprint storehouse of new digit and the weighting calling fingerprint storehouse of old number is calculated, according to the similarity Judge whether the new digit attaches most importance to networking user.
5. the method for identification reentry network users according to claim 4, it is characterised in that effective in new/old number exhales Cry in fingerprint base, the call features weight of each opposite-terminal number is calculated according to equation below：
w u i = T F - IDF u i Σ i = 1 n ( tf u i ) 2 * Σ i = 1 n ( idf u i ) 2 ;
T F - IDF u i = T u i T i * tf u i * idf u i ;
Wherein, wuiIt is i-th call features weight of opposite-terminal number in the effective call fingerprint base of number u；TF-IDFuiIt is number I-th influence coefficient of opposite-terminal number in the effective call fingerprint base of u；N is opposite-terminal number in the effective call fingerprint base of number u Sum；TuiIt is i-th opposite-terminal number and the duration of calls of the number u in setting time；TiIt is that i-th opposite-terminal number is setting With institute's number duration of call summation in time；tfuiIt is i-th opposite-terminal number and calls of the number u in setting time time Number, idfuiFor i-th opposite-terminal number in setting time with institute's number talk times summation.
6. it is according to claim 4 identification reentry network users method, it is characterised in that calculate new number according to equation below The weighting calling fingerprint storehouse of code and the similarity for weighting calling fingerprint storehouse of old number：
S i m ( D 1 , D 2 ) = c o s θ = Σ k = 1 n w 1 k × w 2 k ( Σ k = 1 n w 1 k 2 ) ( Σ k = 1 n w 2 k 2 ) ;
Wherein, Sim (D1,D2) represent number 1 weighting calling fingerprint storehouse D1With the weighting calling fingerprint storehouse D of number 22It is similar Degree；w1kIt is the weighting calling fingerprint storehouse D of number 11In k-th opposite-terminal number call features weight；w2kIt is the weighting of number 2 Calling fingerprint storehouse D2In k-th opposite-terminal number call features weight；N be effective call fingerprint base in opposite-terminal number it is total Number.
7. it is according to claim 4 identification reentry network users method, it is characterised in that according to the similarity judge institute Whether stating the new digit networking user that attaches most importance to includes：
If the similarity is in the number range of setting, determine that the new digit is attached most importance to networking user.
8. it is a kind of recognize reentry network users device, it is characterised in that including：
Calling fingerprint acquisition module, the detailed list information of calling for obtaining new/old number, according to the calling, list information is distinguished in detail Set up the calling fingerprint storehouse of new/old number；Packet in the calling fingerprint storehouse contains：It is all to be conversed with new/old number Opposite-terminal number, and call characteristic information；The call characteristic information of each opposite-terminal number includes：Itself and described new/old number Talk times and the duration of call, and itself and institute's number talk times summation and duration of call summation；
Fingerprint filtering module, for the call characteristic information according to each opposite-terminal number, calculates the calling fingerprint of new/old number respectively The influence coefficient of each opposite-terminal number in storehouse；Coefficient mistake from the calling fingerprint storehouse of new/old number is influenceed according to each opposite-terminal number Infecctive call fingerprint is filtered, effective call fingerprint base is drawn；
Network re-entry determination module, for the phase of the effective call fingerprint base according to new digit and the effective call fingerprint base of old number Like spending, judge whether the new digit attaches most importance to networking user.
9. it is according to claim 8 identification reentry network users device, it is characterised in that the fingerprint filtering module bag Include：
Influence coefficient calculation unit, for each opposite-terminal number in the calling fingerprint storehouse that new/old number is calculated according to equation below Influence coefficient：
T F - IDF u v = T u v T v * tf u v * idf u v ;
Wherein, TF-IDFuvIt is the influence coefficient of opposite-terminal number v in the calling fingerprint storehouse of number u；TuvIt is opposite-terminal number v and number u The duration of call in setting time；TvFor opposite-terminal number v in setting time with institute's number duration of call summation；tfuv It is opposite-terminal number v and talk times of the number u in setting time；idfuvFor opposite-terminal number v in setting time with all numbers The talk times summation of code；
And/or,
Sequence and screening unit, for the opposite-terminal number and its call characteristic information in the calling fingerprint storehouse of new/old number to be pressed It is ranked up from high to low according to influence coefficient, obtains the opposite-terminal number and its call characteristic information of the preceding setting number of sequence, Draw effective call fingerprint base.
10. it is according to claim 8 identification reentry network users device, it is characterised in that the network re-entry determination module Including：
Weighting processing unit, in the effective call fingerprint base of new/old number, the influence coefficient according to each opposite-terminal number, The call features weight of each opposite-terminal number is determined respectively, obtains the weighting calling fingerprint storehouse of the new/old number；
Similarity calculated, the phase for calculating the weighting calling fingerprint storehouse of new digit and the weighting calling fingerprint storehouse of old number Like degree；
And, identifying unit, for judging whether the new digit attaches most importance to networking user according to the similarity.
CN201611117570.4A 2016-12-07 2016-12-07 The method and apparatus for recognizing reentry network users Pending CN106791222A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201611117570.4A CN106791222A (en) 2016-12-07 2016-12-07 The method and apparatus for recognizing reentry network users

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201611117570.4A CN106791222A (en) 2016-12-07 2016-12-07 The method and apparatus for recognizing reentry network users

Publications (1)

Publication Number Publication Date
CN106791222A true CN106791222A (en) 2017-05-31

Family

ID=58881819

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201611117570.4A Pending CN106791222A (en) 2016-12-07 2016-12-07 The method and apparatus for recognizing reentry network users

Country Status (1)

Country Link
CN (1) CN106791222A (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109740675A (en) * 2019-01-07 2019-05-10 上海应用技术大学 A kind of same user's discriminance analysis model based on big data
CN112188474A (en) * 2020-10-21 2021-01-05 中国联合网络通信集团有限公司 Re-network user identification method, platform, computer equipment and storage medium

Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102387512A (en) * 2011-10-19 2012-03-21 北京交通大学 Repeat network access analysis method based on vector space model (VSM)

Patent Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102387512A (en) * 2011-10-19 2012-03-21 北京交通大学 Repeat network access analysis method based on vector space model (VSM)

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
罗亚: "移动电话用户重入网识别及营销建议", 《中国优秀硕士学位论文全文数据库》 *
蒋仕宝，陈少权: "基于呼叫指纹的重入网识别算法研究", 《移动通信》 *

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109740675A (en) * 2019-01-07 2019-05-10 上海应用技术大学 A kind of same user's discriminance analysis model based on big data
CN112188474A (en) * 2020-10-21 2021-01-05 中国联合网络通信集团有限公司 Re-network user identification method, platform, computer equipment and storage medium
CN112188474B (en) * 2020-10-21 2022-08-02 中国联合网络通信集团有限公司 Re-network user identification method, platform, computer equipment and storage medium

Similar Documents

Publication Publication Date Title
WO2020215571A1 (en) Sensitive data identification method and device, storage medium, and computer apparatus
CN108280368B (en) Correlation method of online data and offline data and computing equipment
CN109873812A (en) Method for detecting abnormality, device and computer equipment
CN107026831B (en) Identity recognition method and device
CN105608179B (en) The method and apparatus for determining the relevance of user identifier
TW201937437A (en) Method and device for recognizing suspicious money laundering group
CN110046929B (en) Fraudulent party identification method and device, readable storage medium and terminal equipment
CN110399490A (en) A kind of barrage file classification method, device, equipment and storage medium
CN105306495B (en) user identification method and device
CN111274338B (en) Pre-outbound user identification method based on mobile big data
CN110197402B (en) User label analysis method, device, equipment and storage medium based on user group
WO2022007559A1 (en) Palm print recognition method, feature extraction model training method, device and medium
CN109978020A (en) A kind of social networks account vest identity identification method based on multidimensional characteristic
CN113837323B (en) Training method and device of satisfaction prediction model, electronic equipment and storage medium
CN102387512B (en) Repeat network access analysis method based on vector space model (VSM)
CN107092592A (en) A kind of personalized method for recognizing semantics in the place based on type multiple-situation data and cost-sensitive integrated model
CN110457677A (en) Entity-relationship recognition method and device, storage medium, computer equipment
CN108228684A (en) Training method, device, electronic equipment and the computer storage media of Clustering Model
CN109829302A (en) Android malicious application family classification method, apparatus and electronic equipment
CN112488716A (en) Abnormal event detection system
CN107741958A (en) A kind of data processing method and system
CN107516370A (en) The automatic test and evaluation method of a kind of bank slip recognition
CN106910135A (en) User recommends method and device
CN110046648A (en) The method and device of business classification is carried out based at least one business disaggregated model
CN106791222A (en) The method and apparatus for recognizing reentry network users

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
RJ01 Rejection of invention patent application after publication
RJ01 Rejection of invention patent application after publication

Application publication date: 20170531